Databend Cloud

In this document we'll show how to set up a simple data pipeline to load data from indexed.xyz to Databend Cloud and perform analysis with SQL.

Currently, Databend Cloud is under beta access, and offers a trial with $200 credit on their homepage

Load data from indexed.xyz

Indexed.xyz file sare stored in cloudflare R2, and for this example, we'll be using the prefix 9d for the demo. We can set up a Databend worksheet to run all queries with an XSmall instance.

First, we need to add indexed.xyz R2 as a Databend stage.

CREATE STAGE r2_stage
URL = 's3://indexed-xyz/ethereum/decoded/logs/v1.2.0/partition_key=9d/'
CONNECTION = (
  ENDPOINT_URL = 'https://data.indexed.xyz/'
  REGION = 'auto'
  ACCESS_KEY_ID = '43c31ff797ec2387177cabab6d18f15a'
  SECRET_ACCESS_KEY = 'afb354f05026f2512557922974e9dd2fdb21e5c2f5cbf929b35f0645fb284cf7'
);

You can take a look at the loaded files with the following command:

LIST @r2_stage;

This should return something like the following table:

name	size	md5	last_modified	creator
dt=2015/1680119600-25ee0299-28eb-47a4-b81e-f49f5399617b-0-4.parquet	12616	e273f913a5f75fea96f9fe7d2f086a6f	2023-04-07 02:02:21.717 +0000	NULL
dt=2015/1680119600-25ee0299-28eb-47a4-b81e-f49f5399617b-0-42.parquet	5470	a29c8427372d02dece62924dd985ae51	2023-04-07 02:02:21.725 +0000	NULL
dt=2016/1680119600-25ee0299-28eb-47a4-b81e-f49f5399617b-37-26.parquet	32708	2c21a9ffbec6d900f21fe4e8ff30d00f	2023-04-07 02:02:22.131 +0000	NULL
dt=2016/1680119600-25ee0299-28eb-47a4-b81e-f49f5399617b-37-27.parquet	261112	8efed0a4490b52bb35a12eb1d2162d18	2023-04-07 02:02:21.724 +0000	NULL
dt=2016/1680119600-25ee0299-28eb-47a4-b81e-f49f5399617b-37-44.parquet	10927	d5f7c980ac9e1e1c3de8d42c4388a7b1	2023-04-07 02:02:21.794 +0000	NULL
dt=2016/1680119600-25ee0299-28eb-47a4-b81e-f49f5399617b-37-45.parquet	273489	e41f8268c73ff9e8062a16fd1bd3d8bf	2023-04-07 02:02:22.180 +0000	NULL
dt=2016/1680119600-25ee0299-28eb-47a4-b81e-f49f5399617b-37-9.parquet	4220496	f206d6bcb48483dc684ba8c67d5c6733	2023-04-07 02:02:22.145 +0000	NULL

Create contract table

Build the target table using the following SQL:

CREATE DATABASE indexedxyz;

USE indexedxyz;

CREATE TABLE `contract` (
  `block_time` BIGINT NULL,
  `address` VARCHAR NULL,
  `event_signature` VARCHAR NULL,
  `event_params` ARRAY(STRING NULL) NULL,
  `block_number` BIGINT NULL,
  `block_hash` VARCHAR NULL,
  `log_index` BIGINT NULL,
  `transaction_hash` VARCHAR NULL,
  `transaction_index` BIGINT NULL,
  `data` VARCHAR NULL,
  `topics` VARCHAR NULL,
  `id` VARCHAR NULL
);

Load data from stage

COPY INTO contract FROM @r2_stage FILE_FORMAT = (type = PARQUET);

The above command is idempotent, once files are loaded into the contract table, redoing the copy would not load the file again. Once data is loaded into the contract table (typically takes 1-10 minutes), we can start to query it.

Counting the tokens

We could check on BAYC mints under the partiion 9d by querying the data:

SELECT COUNT(DISTINCT(CAST(event_params[3] AS INTEGER)))
FROM contract
WHERE LOWER(address) = '0xbc4ca0eda7647a8ab7c2061c2e118a18a936f13d'
  AND event_signature = 'Transfer(address,address,uint256)';

| count() | | 5374 |

And if we want to check on the transaction volume by month, we can use the following command:

SELECT
    date_trunc('month', to_timestamp(block_time)),
    count(*)
FROM
    contract
WHERE
    lower(address) = '0xbc4ca0eda7647a8ab7c2061c2e118a18a936f13d'
    AND event_signature = 'Transfer(address,address,uint256)'
GROUP BY
    1;

The result should be:

date_trunc(month, to_timestamp(block_time))	count(*)
2021-05-01	1742
2022-04-01	352
2023-03-01	1622
2022-07-01	244
2023-01-01	2556
2021-12-01	262
2021-10-01	190
2022-06-01	257
2022-02-01	205
2022-08-01	285
2021-07-01	357
2022-10-01	225
2021-11-01	219
2022-11-01	334
2021-08-01	563
2021-06-01	727
2023-04-01	121
2022-12-01	1566
2022-01-01	272
2022-05-01	426
2022-03-01	415
2021-04-01	48
2021-09-01	257
2023-02-01	1950
2022-09-01	245

Visualize results through Databend worksheet

Databend Cloud natively provides a visualization toolkits on worksheets.